MINEing metabolism for unannotated enzymatic functions and serendipitous metabolic pathways

Event Sponsor: 
Mathematics and Computer Science Division Seminar
Start Date: 
May 31 2017 - 10:30am
Building 240/Room 4301
Argonne National Laboratory
James Jeffryes
Chris Henry

The potential of metabolic engineering is greatly enhanced by enzymes’ natural flexibility, which permits them to catalyze the transformation of not only their native substrate but also a range of related substrates. This trait enables cells to produce not only natural metabolic products, but also valuable compounds for which no natural pathway exists through rational design and directed evolution. However, unless physically sequestered, these enzymes and metabolic intermediates continue to interact with the chassis’ preexisting metabolic network. Understanding potential off-pathway effects of overexpressed enzymes or the alternate fates of novel pathway intermediates permits the rationalization of observed phenotypes and informs the design of new pathways.

To address this challenge, we have constructed Metabolic In silico Network Expansions (MINEs) that expand existing metabolic models by using expert curated reaction rules to propose novel metabolites and reactions. The reaction rules include an enzymatic set, which has been demonstrated to reproduce a large fraction of known biochemical reactions and a set describing spontaneous (non-enzymatic) chemical transformations of metabolites under physiological conditions. We have applied these generalized reaction rules to compounds from various biochemical databases such as KEGG (www.kegg.jp) and the Model SEED (www.theseed.org/models) The resulting MINEs are freely accessible for noncommercial use at http://minedatabase.mcs.anl. The databases contain over 750,000 putative metabolites; over 90% of which are not found in PubChem, the largest freely available database of chemicals. MINEs have been used to annotate novel metabolites from 4 diverse organisms and propose potential sources for these compounds within known metabolism. MINE databases shine a light on unannotated enzymatic functions and serendipitous metabolic pathways, enabling more complete and predictive models of cellular metabolism.

Speaker Biography: James Jeffryes focuses on the prediction and modeling of cellular metabolism to understand complex biological systems and annotate metabolomics data. He defended his doctoral dissertation Chemical Engineering at Northwestern University where he served as a Leadership Fellow. James is passionate about the development of open source computational tools and has contributed to RDKit and dask projects as well as developing the MINE databases.